BATTLING MISDIAGNOSIS WITH AI

2nd Opinion is an AI-powered app that helps users compare their symptoms and possible causes with the diagnosis they received from their doctor, aiming to support—not replace—professional medical advice.

The app provides data-driven insights to keep patients informed and assist in managing follow-up care. While it aims for accuracy, precision is not yet guaranteed, and ongoing improvements are being made to enhance its reliability.

The growing demand for medical services, coupled with a shortage of healthcare providers, has strained the system to its limits. This has resulted in issues that negatively affect patient care and the overall efficiency.

1. Misdiagnosis


Diagnostic errors contribute to around 795,000 deaths or cases of permanent disability in the U.S.annually.

2. Mistreatment


Mistreatment often results from misdiagnosis, leading to incorrect treatments that could harm patients, with diagnostic errors affecting up to 20% of medical cases.

3. High cost


As of 2024, U.S. healthcare costs are projected to reach approximately $4.3 trillion annually, reflecting ongoing increases in healthcare spending.

4. Lengthy wait time


Patients in U.S. usually wait up to 70 days to see specialists, leading to delays in care.

5. Lack of doctor’s incentive and accountability


Doctors sometimes lack incentives to double-check diagnoses, leading to errors that impact patient outcomes.

2nd Opinion was developed as an AI-driven app to address these critical issues in the healthcare industry. After researching the market and competitors, we discovered that there was a big gap in providing accurate and affordable health checkups, particularly in the U.S. and Canada. Moreover, recent studies show that AI-based products like ChatGPT and AMIE have outperformed doctors in diagnostic tests.

The iOS 2nd Opinion app is built with a Mixture of Agents, all based on ML models, such as OpenAI’s GPT, WhisperAI for speech recognition, and Llama for medical language processing. These ML models are trained on diverse healthcare datasets, including clinical guidelines, anonymized patient records, and symptom databases.

The app uses SwiftUI with HealthKit integration for a personalized user experience. On the infrastructure side, AWS Amplify provides the backbone, with Cognito handling secure authentication, DynamoDB managing scalable data storage, and Lambda powering serverless operations. This setup ensures the app scales efficiently and integrates new features easily as it grows.

2nd Opinion offers AI-driven health assessments, provides affordable access to health information with a simple monthly subscription, and ensures 24/7 availability.

Medical record


The app has an integration with Apple Health, which transfers medical info, including data from wearable devices, 23andMe, MRIs, and ECGs. This makes the app context-aware and personalized.

Human-like experience


Users can interact with the app as they would with their doctors: by calling, texting, or sending files. One minor point — compared to a doctor, the app is available 24/7.

Separate profiles


Every user owns their profile, but parents can create their child profiles and keep a health track record in one place.

Data privacy & security


Users’ health data is protected by the highest data privacy and security standards, ensuring it remains confidential.
Health reports
2nd Opinion’s advanced AI provides health insight through an extensive medical database. Users can generate detailed health reports to share with their healthcare providers for further consultation.

• Integrated the app with Apple Health, so now it offers a human-like, proactive healthcare experience.
• Implemented a Mixture of Agents built on advanced ML models.
• Validated the market.

Focused on enhancing the integration and application of the Mixture of Agents to achieve improved accuracy and reliability, while also validating the monetization potential.